steipete/clawdis
102 skills188k installs39M starsGitHub
Install
npx skills add https://github.com/steipete/clawdisSkills in this repo
1Summarizesummarize is a fast CLI tool for extracting and summarizing content from URLs, local files, YouTube videos, and podcasts. Developers use it when they need to quickly process long-form content into concise summaries. It supports multiple LLM providers (OpenAI, Anthropic, xAI, Google) with configurable summary lengths and fallback extraction via Firecrawl.22.3kinstalls2WeatherThis skill retrieves current weather data and multi-day forecasts from wttr.in for any city, region, airport code, or coordinates. Developers use it when building agents or workflows that need real-time weather context for travel planning, rain checks, or temperature queries. The primary method is web_fetch with JSON format (j2) to avoid HTML responses from browser-like User-Agents, parsing fields like temp_C, weatherDesc, precipMM, and forecast dates from the response. A curl fallback is provided for environments where web_fetch is unavailable, supporting multiple wttr.in format strings. The skill also documents reliability fallbacks to wttr.is and warns against using wttr.in for severe alerts, historical climate data, or aviation decisions.6.4kinstalls3Tmuxtmux is a terminal multiplexer control skill for agents and scripts managing interactive CLI sessions. Developers use it to automate workflows that require real-time interaction with long-running processes, capturing pane output, sending keys/text, and monitoring prompts. Key workflows include listing sessions/windows/panes, capturing full scrollback history, sending literal text or special keys (Ctrl+C, Ctrl+D), and waiting for specific prompt patterns before responding. Critical for testing scenarios where agent tooling must interact with interactive CLIs that don't support non-interactive flags, and for monitoring multi-pane setups that persist across SSH disconnects. --- name: tmux description: "Control tmux sessions/panes for interactive CLIs: list, capture output, send keys, paste text, monitor prompts." metadata: { "openclaw": { "emoji": "🧵", "os": ["darwin", "linux"], "requires": { "bins": ["tmux"] }, "install": [ { "id": "brew", "kind": "brew", "formula": "tmux", "bins": ["tmux"], "label": "Install tmux (brew)", }, ], }, } --- # tmux Use for existing interactive tmux sessions.5.8kinstalls4Goggog is the Google Workspace CLI at gogcli.sh for Gmail, Calendar, Drive, Contacts, Sheets, and Docs automation from scripts and agent shells. One-time OAuth setup runs gog auth credentials with client_secret.json, gog auth add per account selecting gmail calendar drive contacts docs sheets services, and gog auth list for verification. Gmail covers thread search, per-message search ignoring threading, send with plain body, body-file, stdin heredocs, or HTML, drafts create and send, and reply with reply-to-message-id. Calendar lists and creates events with optional --event-color from IDs 1 through 11 via gog calendar colors. Drive search, contacts list, sheets get update append clear metadata, and docs export or cat are documented with --json for scripting. Email guidance prefers plain text via --body-file because --body does not unescape newlines; HTML uses --body-html with documented tags. Set GOG_ACCOUNT to avoid repeating --account; use --json and --no-input in scripts. Confirm before sending mail or creating events.5.1kinstalls5Nano PdfThe nano-pdf skill Edit PDFs with natural-language instructions using the nano-pdf CLI.. nano-pdf Use `nano-pdf` to apply edits to a specific page in a PDF using a natural-language instruction. Quick start Notes: - Page numbers are 0-based or 1-based depending on the tool's version/config; if the result looks off by one, retry with the other. - Always sanity-check the output PDF before sending it out. Agents should read SKILL.md quick start steps, verify required binaries and environment variables, and follow reference files for exact parameters before calling tools. The workflow includes decision gates, anti-patterns, and cross-links to sibling skills so agents stay aligned with upstream documentation rather than improvising. Use when user language matches documented trigger phrases or when the task clearly fits the capability described in the skill description field.4.6kinstalls6Openai WhisperThe openai-whisper skill Local speech-to-text with the Whisper CLI (no API key).. Whisper (CLI) Use `whisper` to transcribe audio locally. Quick start - `whisper /path/audio.mp3 --model medium --output_format txt --output_dir .` - `whisper /path/audio.m4a --task translate --output_format srt` Notes - Models download to `~/.cache/whisper` on first run. - `--model` defaults to `turbo` on this install. - Use smaller models for speed, larger for accuracy. Agents should read SKILL.md quick start steps, verify required binaries and environment variables, and follow reference files for exact parameters before calling tools. The workflow includes decision gates, anti-patterns, and cross-links to sibling skills so agents stay aligned with upstream documentation rather than improvising. Use when user language matches documented trigger phrases or when the task clearly fits the capability described in the skill description field.4.6kinstalls7Ordercliordercli is a command-line tool for querying and managing food delivery orders from Foodora (with Deliveroo support in progress). Developers and power users invoke it to retrieve order history, monitor active deliveries with real-time status updates, and programmatically reorder previous meals. Core workflows include listing active orders with `ordercli foodora orders`, viewing detailed order history with `ordercli foodora history`, tracking delivery progress using the `--watch` flag, and initiating reorders via `ordercli foodora reorder`. Authentication supports password login, browser-based login for Cloudflare/bot protection, and session import from Chrome cookies without re-entering credentials. The tool handles multi-country Foodora deployments through configuration commands and outputs structured JSON for integration into automation pipelines. List active orders with arrival status and watch real-time delivery updates using --watch flag. Query order history with optional JSON output for scripting and integration.4.1kinstalls81password1password is a security integration skill for the 1Password op CLI bundled in steipete/clawdis. It walks agents through verifying op --version, detecting auth mode among service account tokens, desktop app integration, or standalone signin, and confirming access with op whoami before any secret read. Service accounts use OP_SERVICE_ACCOUNT_TOKEN for headless direct exec, while desktop integration runs op directly without tmux so IPC to the 1Password app remains reachable on macOS, Linux, and Windows. Standalone signin uses a persistent tmux session with eval-wrapped op signin only when neither service account nor desktop integration is available. Common operations include op read for vault secrets, op run to execute commands with injected environment variables from op:// references, and op inject to render templates into config files. Guardrails forbid pasting secrets into logs or chat and prefer op run and op inject over writing credentials to disk. Developers reach for 1password when scripts or agents need database passwords, API tokens, or SSH keys without exposing them in shell history or plaintext env files.3.9kinstalls9McporterThe mcporter skill List, configure, authenticate, call, and inspect MCP servers/tools with mcporter over HTTP or stdio. Use mcporter to work with MCP servers directly. - mcporter list - mcporter list <server --schema - mcporter call <server.tool key=value - Selector: mcporter call linear.list_issues team=ENG limit:5 - Function syntax: mcporter call "linear.create_issue(title: \"Bug\")" - Full URL: mcporter call https://api.example.com/mcp.fetch url:https://example.com - Stdio: mcporter call --stdio "bun run ./server.ts" scrape url=https://example.com - JSON payload: mcporter call <server.tool --args '{"limit":5}' - OAuth: mcporter auth <server url [--reset] - Config: mcporter config list get add remove import login logout - OAuth: mcporter auth <server url [--reset] - Config: mcporter config list get add remove import login logout - mcporter daemon start status stop restart - CLI: mcporter generate-cli --server <name or --command <url - Inspect: mcporter inspect-cli <path [--json] - TS: mcporter emit-ts <server --mode client types - Config default: ./config/mcporter.json (override with --config). - Prefer --output json for machine-readable results.3.9kinstalls10PeekabooPeekaboo is a macOS UI automation CLI that captures screens, inspects UI elements, drives input, and manages apps, windows, and menus with JSON scriptable output. Commands span bridge connectivity, capture and image screenshots, see for annotated UI maps with snapshot IDs, click, drag, type, scroll, swipe, hotkey, and press interactions, plus app launch, window focus, dock, menu, menubar, dialog, clipboard, and space management. Global flags include --json, --verbose, --bridge-socket, and --no-remote for routing through OpenClaw Bridge so subprocesses reuse Screen Recording and Accessibility grants from the host app socket at ~/Library/Application Support/OpenClaw/bridge.sock. The recommended flow runs peekaboo see --annotate to identify element IDs, then click --on and type with app or window targeting parameters. Capture supports retina screenshots, live region recording with motion-aware FPS, and optional analyze prompts on images. Installation is via brew formula steipete/tap/peekaboo and requires Screen Recording plus Accessibility permissions; macOS 15 adds a separate private window picker grant. The skill excludes agent and MCP features and documents diagnose steps when Bri.3.8kinstalls11Apple Remindersapple-reminders is an agent skill for managing Apple Reminders through the remindctl terminal CLI on macOS. It covers viewing today, tomorrow, week, overdue, or date-filtered reminders, creating and deleting lists, adding items with optional due dates, and completing or deleting reminders by ID. Output formats include JSON for scripting, plain TSV, and quiet counts. Setup requires Homebrew install of steipete/tap/remindctl, Reminders permission, and authorize via remindctl status and remindctl authorize. The skill distinguishes Apple Reminders sync to iPhone and iPad from OpenClaw cron alerts, calendar events, and project task tools like Notion or GitHub Issues. When users say remind me, the agent should clarify whether they want a Reminders app entry or an in-session alert. Accepted due date formats include today, tomorrow, YYYY-MM-DD, and ISO 8601 timestamps for scheduled personal tasks.3.7kinstalls12ObsidianObsidian is an agent skill that drives the official obsidian command-line interface for vault work on a running Obsidian desktop app. It documents requirements for Obsidian 1.12.7 or newer, CLI registration on PATH, and vault resolution via file or path parameters. Covered operations span open and read, JSON search, create and append, move and delete, daily notes, task completion, property get and set, backlinks, unresolved links, plugin reload, and developer eval helpers. The skill distinguishes file-based resolution from exact vault-relative paths, prefers CLI commands for Obsidian-aware moves and property updates, and allows direct Markdown edits for bulk text when the vault path is known. macOS registration creates /usr/local/bin/obsidian; Linux copies the binary to ~/.local/bin/obsidian. Developers reach for it when automating note search, task triage, property updates, or vault maintenance without third-party obsidian-cli wrappers.3.7kinstalls13Coding Agentcoding-agent delegates substantial coding work to Codex, Claude Code, or OpenCode as background workers instead of handling simple edits inline. Use it for feature builds, PR reviews, large refactors, and issue-to-PR loops, but not for trivial edits, read-only lookup, ACP thread-bound work, or runs inside OpenClaw state directories. Hard rules require background:true launches, pty:true for Codex and OpenCode, Claude Code with --permission-mode bypassPermissions --print without PTY, a real notification route before spawn, and completion via openclaw message send rather than heartbeat or notify-on-exit. Workers receive prompts written to temp files to avoid shell quoting bugs, including a notification block with channel, target, account, reply_to, and thread_id fields. Launch forms document codex exec, claude --print, and opencode run commands with workdir set to an isolated repo checkout, never ~/Projects/openclaw. Long issue-to-PR flows create GitHub issues as durable specs, branch, test, review until clean, and open PRs while returning sessionId immediately. Process actions cover list, poll, log, submit, write, paste, and kill for monitoring slow workers.3.7kinstalls14Himalayahimalaya teaches agents to operate the Himalaya CLI for IMAP and SMTP email from the terminal. Setup verifies himalaya --version, runs himalaya account configure, and stores config at ~/.config/himalaya/config.toml with credentials kept in password managers rather than chat logs. Read and search commands cover folder list, envelope list, message read, and filtered envelope queries by sender or subject. Write workflows use message write, template write, template send from files, reply, and forward, with MML syntax documented in references/message-composition.md for attachments and rich messages. Organization commands copy, move, delete messages and add or remove flags such as seen. Safety rules require confirmation before sending, deleting, or bulk moves, --account when multiple accounts exist, and quoting exact message IDs in summaries. Reference files configuration.md and message-composition.md cover account auth, backend setup, and MML compose syntax. Install metadata supports Homebrew formula himalaya when the binary is missing.3.6kinstalls15ClawhubClawHub teaches OpenClaw agents to discover, install, verify, update, and publish skills from the public registry at clawhub.ai. Discovery starts with openclaw skills search before claiming a capability is unavailable. Before installation the agent must verify the selected skill with openclaw skills verify, treat third-party packages as untrusted, and obtain user approval. Install supports pinned versions and --global for shared local agent installs visible unless allowlists narrow scope. Maintenance commands cover openclaw skills list, check, update for one skill or --all. Publishing workflows use the standalone clawhub CLI after npm i -g clawhub, clawhub login, clawhub whoami, then clawhub skill publish or clawhub sync --all for publisher accounts. The skill separates day-to-day openclaw skills management from clawhub publisher operations. Default installs land in the active workspace while global installs share skills across local agents on the machine.3.6kinstalls16Apple Notesapple-notes is a macOS-only agent skill that wraps the memo CLI for terminal control of Apple Notes. Setup installs memo via Homebrew tap antoniorodr/memo or pip after cloning, and requires Automation access to Notes.app when prompted. View commands list all notes, filter by folder with -f, and fuzzy-search with -s. Create flows use memo notes -a for interactive editor entry or memo notes -a "Note Title" for quick titled notes. Edit and delete use interactive -e and -d selectors. Move notes between folders with memo notes -m. Export selected notes to HTML or Markdown via memo notes -ex, using Mistune for markdown processing. Documented limitations include no editing of notes with images or attachments and interactive prompts that may need terminal access. The skill targets developers automating note capture, folder organization, and export from Apple Notes without opening the GUI on darwin hosts with Notes.app Automation permission granted.3.5kinstalls17Video Framesvideo-frames is a steipete/clawdis skill that extracts still images from video files using ffmpeg through a bundled frame.sh wrapper script. It requires the ffmpeg binary on PATH and documents brew-based installation via formula ffmpeg in OpenClaw metadata. The quick start covers extracting the first frame with --out to a jpg or png path, or targeting a timestamp with --time HH:MM:SS for context around a specific moment. Developers reach for video-frames when an agent needs frame grabs for README screenshots, UI inspection, or lightweight social assets without hand-writing ffmpeg flags. The skill recommends jpg for quick sharing and png for crisp UI frames. Notes emphasize preferring --time when the question is what is happening around a point in the clip. It fits media-heavy agent workflows where video review must become inspectable stills on disk.3.5kinstalls18Session Logssession-logs is an OpenClaw agent skill for searching and analyzing complete conversation history stored as session JSONL files using jq and ripgrep. Session logs live under OPENCLAW_STATE_DIR agents agentId sessions with sessions.json mapping chat providers to session ids and per-session jsonl transcript files. Each line records type session or message, ISO timestamps, message roles user assistant or toolResult, message content text thinking or tool calls, and message usage cost totals for billing summaries. Documented queries list sessions by date and size, filter sessions by day, extract user or assistant text, search assistant responses for keywords, sum per-session and daily costs, count messages and roles, rank tool call frequency, and ripgrep across all sessions for phrases. Tips note append-only large files where head and tail help sampling and deleted sessions carry a deleted timestamp suffix. Developers reach for it when users reference parent conversations, prior chat context missing from memory files, or historical cost and tool usage analysis across OpenClaw agent sessions.3.3kinstalls19Trellotrello is an OpenClaw skill for managing Trello boards, lists, and cards via the Trello REST API. Setup requires an API key from trello.com/app-key, a generated token, and TRELLO_API_KEY plus TRELLO_TOKEN environment variables that grant full account access and must stay secret. Commands use curl with key and token query parameters piped through jq to list boards, list board lists, list cards, create cards with idList name and desc, move cards by updating idList, add comments, and archive cards with closed=true. Examples filter boards by name and list all cards on a board with list ids. Notes cite rate limits of 300 requests per 10 seconds per API key and 100 per 10 seconds per token, with /1/members endpoints limited to 100 requests per 900 seconds. Metadata requires curl and jq binaries. Developers reach for it when an agent must read or mutate Trello boards without a separate SDK.3.3kinstalls20Healthcheckhealthcheck is an agent skill for assessing and hardening machines running OpenClaw. It first infers OS, container vs host, privilege level, access path, network exposure, gateway bind and auth, backups, disk encryption, automatic updates, and usage mode, asking only for missing facts. After permission, it runs read-only checks such as openclaw security audit --deep, openclaw gateway status --deep, openclaw doctor, plus platform commands for listening ports, firewall state, SSH status, encryption, and scheduled updates on macOS, Linux, or Windows. The user then picks a risk posture from convenience, balanced, or strict. Reports summarize current posture, severity-ranked findings with evidence, a staged reversible plan, read-only commands first, and gaps where checks could not run. Hardening menu items cover gateway bind targets, remote auth, closing public ports, OS updates, disk encryption, backup verification, SSH key requirements, and scheduled deep audits. Rules require asking before state-changing actions, never claiming OpenClaw manages OS firewall or SSH, never printing secrets, and confirming exact actions before applying changes.3.3kinstalls21Githubgithub is an agent skill for the GitHub CLI covering authentication, pull requests, issues, CI workflow runs, releases, and gh api queries. It directs agents to use gh for GitHub operations, git for local commits and branches, and code-reading tools for deep reviews. Auth commands include gh auth status and gh auth login, with GH_CONFIG_DIR guidance when gateway HOME differs from operator HOME. Pull request recipes list, view JSON metadata, inspect checks and diffs, create from body files, and squash merge. Issue workflows cover list, view, create, comment, and close with body-file patterns for backtick-heavy content. CI commands list runs, view status and failed logs, and rerun failed jobs. API examples fetch pull request fields, labels, and cached repository star and fork counts with jq filtering. The skill emphasizes --json plus --jq for structured output and --body-file for comments containing shell snippets or user text. Developers reach for it when automating PR review, issue triage, CI debugging, or GitHub API lookups from agent environments requiring gh.3.2kinstalls22BlogwatcherBlogwatcher is a lightweight CLI skill for monitoring blogs and RSS or Atom feeds through the blogwatcher binary. Install via go install github.com/Hyaxia/blogwatcher/cmd/blogwatcher@latest as documented in OpenClaw metadata requiring the blogwatcher bin. Quick start begins with blogwatcher --help. Core commands add a blog with blogwatcher add, list tracked blogs via blogwatcher blogs, scan for updates with blogwatcher scan, list articles with blogwatcher articles, mark one article read with blogwatcher read, mark all read with blogwatcher read-all, and remove a blog with blogwatcher remove. Example scan output reports per-blog source type RSS, found count, and new article count, then totals new articles across all tracked feeds. The skill is intentionally minimal: agents discover additional flags through per-command --help. Typical workflow adds feed URLs, periodically scans, surfaces new articles, and marks items read to keep an unread queue manageable for content monitoring or research digests.3.2kinstalls23SonoscliSonos CLI is a skill for the sonos binary that controls Sonos speakers via local network discovery and named targeting. Install with go install github.com/steipete/sonoscli/cmd/sonos@latest per OpenClaw metadata. Quick commands include sonos discover, sonos status --name, play pause stop, and sonos volume set with a speaker name. Grouping covers group status, join, unjoin, and party modes. Favorites support list and open, queue supports list play and clear, and Spotify search runs via SMAPI with sonos smapi search --service Spotify. If SSDP discovery fails, pass --ip for a known speaker address. Optional Spotify Web API search needs SPOTIFY_CLIENT_ID and SPOTIFY_CLIENT_SECRET environment variables. Troubleshooting documents no route to host on macOS direct mode requiring Local Network permission for the Gateway parent process such as node launchd, Terminal, or VS Code, or using Docker sandbox with network allowed. Bind operation not permitted errors indicate Codex or other sandboxes blocking UDP; users must approve network escalation or run outside the sandbox.3.2kinstalls24Model UsageModel Usage is a macOS-oriented skill for reading per-model AI usage costs from CodexBar local cost logs. It supports current model mode using the most recent daily row with modelBreakdowns and picking the highest-cost model, or all models mode for full summaries. Providers include codex and claude. Default input runs codexbar cost --format json --provider, while file or stdin input via --input works on any platform with exported JSON. The bundled scripts/model_usage.py handles summarization with optional --model override, text or JSON --format output, and --pretty printing. Current model logic falls back to the last entry in modelsUsed when breakdowns are missing. CodexBar CLI live reads are documented for macOS with brew cask install steipete/tap/codexbar; Linux and other platforms should pass exported JSON until a supported local install path exists. Output values are cost-only per model because CodexBar does not split tokens by model. References/codexbar-cli.md documents CLI flags and cost JSON fields for deeper inspection.3.2kinstalls25Spotify Playerspotify-player is a steipete/clawdis agent skill for terminal Spotify playback and search via the spogo CLI, with spotify_player as fallback when spogo is unavailable. It requires a Spotify Premium account and at least one of the two binaries installed through Homebrew formulas spogo or spotify_player. Setup imports browser cookies with spogo auth import --browser chrome so the CLI can authenticate without manual token pasting. Documented spogo commands cover track search, play pause next prev transport controls, device listing and selection, and status queries. The spotify_player fallback exposes search, playback, connect, and like commands with configuration stored under ~/.config/spotify-player including optional client_id for Spotify Connect. OpenClaw metadata declares required bins and brew install recipes so agents can verify prerequisites before attempting playback. Developers and agents invoke it when users want hands-free Spotify control from a coding session, shell automation, or OpenClaw bridge workflows without opening the desktop app. The skill stays narrow: it documents CLI syntax and install paths rather than building a custom Spotify API integration.3.2kinstalls26NotionThe Notion skill guides agents through the official ntn CLI for pages, databases, data sources, files, comments, search, Workers, and raw API calls. Setup covers npm global install, interactive ntn login, or headless NOTION_API_TOKEN with NOTION_API_VERSION 2026-03-11. Markdown-first page helpers include pages get with frontmatter properties, pages create and update from content strings or stdin, and pages trash with confirmation. Data source commands resolve database IDs, query with filters and sorts, and return JSON for agent parsing. Raw API access uses path=value body fields, typed path:=json values, query parameters, and schema or docs inspection before constructing payloads. File uploads support stdin binary, external URLs, and upload listing. Workers commands cover new, deploy, list, and run logs where the workspace plan allows. A curl fallback documents Authorization and Notion-Version headers when ntn is unavailable.3.2kinstalls27Openai Whisper ApiThe openai-whisper-api skill wraps OpenAI audio transcriptions through a bundled transcribe.sh script and curl. Default model is gpt-4o-transcribe with text output saved beside the input file as a sibling .txt path. Supported flags switch to gpt-4o-mini-transcribe for lower cost, gpt-4o-transcribe-diarize with JSON speaker labels, legacy whisper-1, explicit language codes, prompt hints for names or jargon, and custom --out paths for JSON transcripts. Upload formats include mp3, mp4, mpeg, mpga, m4a, wav, and webm with a documented 25 MB hosted API limit. OPENAI_API_KEY is required from the environment or OpenClaw skills config, and OPENAI_BASE_URL supports compatible proxies or local gateways. Diarize mode sends chunking_strategy auto and rejects prompt flags per API rules. Install metadata lists curl and node binaries with brew fallback guidance. The skill targets OpenClaw and agent environments that need quick speech-to-text without hand-building multipart HTTP requests.3.1kinstalls28Sherpa Onnx Ttssherpa-onnx-tts is an OpenClaw agent skill for local offline text-to-speech using the sherpa-onnx CLI wrapper and Piper voice models. Install steps download platform-specific runtime archives for macOS Linux and Windows into the OpenClaw tools directory, then fetch a voice model such as vits-piper-en_US-lessac-high into models. Configuration writes SHERPA_ONNX_RUNTIME_DIR and SHERPA_ONNX_MODEL_DIR into the active openclaw.json skills.entries env block after resolving OPENCLAW_STATE_DIR or default ~/.openclaw paths. Usage runs the bin wrapper with -o output.wav and quoted text, with optional SHERPA_ONNX_MODEL_FILE or --model-file when multiple onnx files exist. Windows invocation uses node on the wrapper path. The skill supports darwin linux and win32, requires those environment variables, and notes picking alternate voices from sherpa-onnx tts-models releases. No cloud API is needed once runtime and model directories are present, making it suitable for private agent workflows that need spoken output without network TTS services.3.1kinstalls29Things MacThings Mac is a macOS-only agent skill for the things3-cli binary that reads and writes Cultured Code Things 3 tasks. Read-only commands query the local Things database for inbox, today, upcoming, search, projects, areas, and tags after granting Full Disk Access to Terminal or OpenClaw.app. Write operations use the Things URL scheme to add todos with notes, schedules, deadlines, projects, headings, tags, and checklist items, or to update titles, notes, lists, tags, and completion state when THINGS_AUTH_TOKEN is set. Dry-run mode previews URLs without opening Things. The skill documents install via go install on Apple Silicon, optional THINGSDB overrides, STDIN multi-line title and notes input, and that delete or trash is not supported so completion or cancel is the workaround. It targets agents helping users capture tasks, search existing items, and modify todos from chat without leaving the CLI workflow.3.1kinstalls30Goplacesgoplaces is an agent skill for the goplaces CLI wrapping the modern Google Places API with human-readable output and JSON mode for scripts. Install via Homebrew steipete/tap/goplaces and set GOOGLE_PLACES_API_KEY before running searches, resolves, or detail lookups. Commands cover text search with open-now and rating filters, geo bias with lat-lng-radius, pagination tokens, place resolution from free text, and detail fetches with optional reviews. JSON output uses the --json flag for piping into jq workflows alongside human tables by default. Notes document price level ranges from free to very expensive, single type filter limits, NO_COLOR support, and optional GOOGLE_PLACES_BASE_URL overrides for testing or proxying. OpenClaw metadata documents brew install requirements and primary environment variable naming. Use it when agents need nearby place discovery, address resolution, or structured place metadata inside automation without hand-rolling HTTP calls to Google Places.3.1kinstalls31Openhueopenhue teaches agents to drive Philips Hue lights and scenes with the OpenHue CLI against a local Hue Bridge. It covers listing lights, rooms, and scenes, then turning devices on or off, setting brightness from 0 to 100, adjusting color temperature in mirek, and applying named or hex colors on capable bulbs. Room-level commands dim or shut down entire zones, while scene activation pairs a scene name with a target room such as Relax or Concentrate. Quick presets document bedtime, work mode, and movie mode brightness and temperature combos. The skill states clear boundaries: Hue devices only, not HomeKit, TVs, thermostats, or non-Hue smart plugs. Setup requires the bridge on the local network and a one-time button press to pair on first run. Install path is Homebrew via openhue/cli/openhue-cli. Agents reach for it when users ask to turn lights on, dim a room, or activate a scene by name.3.1kinstalls32OracleOracle is an agent skill for the oracle CLI that bundles a prompt plus selected files into one second-model advisory pass. Default model is gpt-5.5-pro with browser engine for long ChatGPT Pro runs or API engine when OPENAI_API_KEY or Azure config is ready. Golden path picks a tight file set, previews payload with dry-run summary and files-report, then runs browser or API mode without re-running on detach; reattach stored sessions instead. File attachment supports globs, excludes, gitignore honoring, and one megabyte size limits with default ignored directories. Sessions persist under ~/.oracle/sessions with status and attach commands. Prompt guidance stresses zero project knowledge assumptions, stack briefing, repro steps, verbatim errors, constraints, and desired output format. Safety rules forbid attaching secrets from .env or key files. Install via npm @steipete/oracle. Use it for design review, debug second opinions, refactors, or long investigations you may need to rerun days later.3.1kinstalls33Gifgrepgifgrep is an agent skill for the gifgrep CLI that searches GIF providers, previews results, downloads clips, and extracts still frames or multi-frame sheets. Workflow runs search, optional TUI browse, download to ~/Downloads, and still or sheet extraction for docs, PRs, or chat sharing. Commands support Tenor and Giphy sources with API keys, JSON output, thumbnail previews on Kitty and Ghostty terminals, and provider auto selection. Sheet extraction samples frames into a single PNG grid with tunable frames, columns, and padding. Asset hygiene rules require verifying GIF URLs resolve, checking Content-Type image/gif, confirming animation, and recording attribution instead of hotlinking when local copies are needed. Install via Homebrew steipete/tap/gifgrep or go install. Use it when agents need GIF discovery, quick previews, local downloads, or frame grabs for documentation and social content.3kinstalls34Gh Issuesgh-issues automates issue-to-PR workflows using the gh CLI with optional fork mode, watch loops, and review-only passes. Phase 1 resolves the repo from git remote, loads GH_TOKEN from OpenClaw config if needed, and verifies auth. Phase 2 fetches issues with label, milestone, assignee, state, and limit filters. Phase 3 skips candidates with open PRs, existing fix/issue branches, or active local claims stored in gh-issues JSON with two-hour expiry. Phase 4 confirms selection unless --yes or --cron is set, then configures fork remotes. Phase 5 spawns up to eight background workers with prompts requiring minimal fixes, tests, conventional commits, and PR bodies citing the issue. Phase 6 polls workers and reports PR URLs or failures. Reviews-only mode discovers fix/issue PRs, fetches actionable review comments, and spawns one worker per PR. Watch mode loops fetch, spawn, review processing, and sleep on an interval. Flags include --dry-run, --fork, --reviews-only, --cron, and --model for workers.3kinstalls35Skill CreatorSkill Creator guides agents through creating, editing, auditing, and restructuring AgentSkills with compact SKILL.md files. Skills are triggerable workflows where metadata stays visible, the body loads after trigger, and references, scripts, and assets load only as needed. Hard rules keep SKILL.md lean, put trigger-critical facts in quoted frontmatter description, require name and description fields, prefer noun-phrase descriptions with short generic triggers, and move long examples to references/, scripts to scripts/, and templates to assets/. The recommended shape is skill-name/SKILL.md plus optional scripts, references, assets, and agents folders. The edit workflow reads existing skills, removes generic advice the base model already knows, keeps brittle command syntax and safety rules, relaxes prose, and validates YAML frontmatter after edits. Validation runs quick_validate.py and a YAML safe_load loop across skills. No extra README or changelog files belong inside a skill unless they are actual task references. Use when creating new skills, tightening bloated SKILL.md files, or validating frontmatter across a skill repo.3kinstalls36Xurlxurl is an agent skill for the xurl CLI that wraps authenticated X API v2 work with shortcut commands returning JSON. It covers posting, replies, quotes, deletes, reads, search, timelines, mentions, likes, reposts, bookmarks, followers, blocks, mutes, and direct messages, plus media upload with processing status polling before attaching media-id to posts. Auth guidance stresses never reading or printing ~/.xurl, avoiding inline secrets, skipping --verbose in agent sessions, and checking xurl auth status before calls. Raw mode supports arbitrary v2 endpoints when shortcuts are insufficient, and per-request app or oauth2 overrides are documented for multi-account setups. Agents install via Homebrew xdevplatform/tap/xurl or npm @xdevplatform/xurl. Error handling notes 401 and 403 auth mismatches, 429 rate limits with backoff, and media upload failures from file type or processing state. Use it when automating X posts, searching recent tweets, managing social interactions, or calling niche v2 endpoints from agent workflows.3kinstalls37Sagsag is an agent skill for the sag CLI that provides mac-style say UX on top of ElevenLabs text-to-speech. It requires ELEVENLABS_API_KEY or SAG_API_KEY and supports quick commands such as sag "Hello there", sag speak with a named voice, sag voices, and sag prompting for model tips. Default model is eleven_v3 for expressive output, with eleven_multilingual_v2 for stability and eleven_flash_v2_5 for speed. Pronunciation guidance covers respelling, --normalize auto for numbers and URLs, --lang bias, v3 pause tags like [short pause], and v2 SSML break elements where supported. Voice replies for chat agents can write MP3 files with sag -v Clawd -o /tmp/voice-reply.mp3 and return MEDIA paths. Agents should confirm voice and speaker before long output and pick delivery tags such as [whispers] or [excited] for character. Install via brew steipete/tap/sag. Use when users request spoken responses, voice character delivery, or local TTS playback in OpenClaw style agent workflows.3kinstalls38GeminiThe gemini skill wraps Google's Gemini CLI in headless one-shot mode for agents using the gemini binary installed via brew formula gemini-cli on macOS OpenClaw hosts. Positional arguments start interactive mode, so agents should use -p or --prompt for non-interactive runs, optionally with -m model selection and --output-format json for structured responses. Stdin can append to a prompt, enabling piped summaries from files such as notes.md through shell pipes. Extension management commands list, install, and manage gemini extensions, skills, hooks, and MCP integrations through documented subcommands. Notes require one-time interactive auth if needed and explicitly warn against --yolo for safety. OpenClaw metadata specifies darwin OS, required gemini bin, and brew install instructions with emoji and homepage links. Typical uses include quick question answering, JSON extraction, extension discovery, and routing to Gemini skills or MCP tooling without launching a full interactive terminal session.3kinstalls39Bear NotesBear Notes integrates with the Bear app on macOS through the grizzly CLI for creating, reading, searching, and tagging notes. Requirements include Bear installed and running, with grizzly installable via Go module github.com/tylerwince/grizzly/cmd/grizzly. Token-requiring operations such as add-text, tags, and open-note --selected need a Bear API token from Help, API Token, saved to ~/.config/grizzly/token. Common commands pipe content into grizzly create with title and tag flags, open-note with enable-callback and json for reads, add-text with append mode, tags listing, and open-tag search by tag name. Configuration resolves from CLI flags, environment variables like GRIZZLY_TOKEN_FILE, local .grizzly.toml, or ~/.config/grizzly/config.toml with callback URL and timeout defaults. Dry-run and print-url flags preview x-callback-url actions without executing. Bear must stay running and note IDs come from Bear callbacks when reading data back into scripts or agents.3kinstalls40CamsnapCamsnap captures frames or clips from RTSP and ONVIF cameras through the camsnap CLI installable via Homebrew formula steipete/tap/camsnap. Configuration lives at ~/.config/camsnap/config.yaml with cameras added via camsnap add name, host, user, and pass flags. Common commands discover cameras with discover --info, snap still images to an output path, clip short MP4 segments with duration flags, watch for motion above a threshold with an action command, and doctor --probe for connectivity checks. ffmpeg must be on PATH for media capture operations. The skill recommends a short test snap before longer clips to validate camera credentials and stream stability. OpenClaw metadata marks brew install with required camsnap binary and links homepage camsnap.ai for setup context when wiring agents to local security cameras. Motion watch can trigger shell actions when movement exceeds a configured threshold, making it useful for lightweight alerting workflows tied to local cameras.3kinstalls41Songseesongsee is a CLI for generating spectrograms and feature-panel visualizations from audio files or stdin. A basic spectrogram runs with songsee track.mp3. Multi-panel output combines viz types such as spectrogram, mel, chroma, hpss, selfsim, loudness, tempogram, mfcc, and flux via comma-separated or repeated --viz flags. Time slices use --start and --duration with custom -o output paths. Style palettes include classic, magma, inferno, viridis, and gray. Width, height, window, hop, min-freq, and max-freq control FFT sizing and frequency range. Output format accepts jpg or png via --format. WAV and MP3 decode natively; other formats use ffmpeg when available. Multiple viz flags render a combined grid image. Stdin piping supports scripted workflows such as cat track.mp3 piped to songsee. Install via Homebrew steipete/tap/songsee when the binary is missing from PATH on macOS. Output formats include JPG and PNG with configurable dimensions.3kinstalls42Blucliblucli exposes the blu CLI for discovering and controlling Bluesound and NAD network players from scripts or agent workflows. Quick start runs blu devices to pick a target, blu --device id status for playback state, then play, pause, stop, or volume set on the chosen unit. Target selection follows --device id name or alias, the BLU_DEVICE environment variable, then an optional config default in strict priority order. Grouping commands include group status, add, and remove for multi-room layouts. TuneIn integration supports tunein search and tunein play queries for internet radio streams. Automation should pass --json for machine-readable status and transport fields. Install the blu binary with go install github.com/steipete/blucli/cmd/blu@latest when openclaw metadata reports it missing. Operators must confirm the target device before changing playback or volume so the wrong room speaker is not adjusted unintentionally.3kinstalls43Eightctleightctl controls Eight Sleep pods for status, power, temperature, alarms, schedules, audio, and adjustable base settings from the command line. Authentication uses ~/.config/eightctl/config.yaml or EIGHTCTL_EMAIL and EIGHTCTL_PASSWORD environment variables before any mutating command. Quick commands include eightctl status, on, off, and temp with a celsius target for nightly comfort changes. Alarm workflows support list, create, and dismiss operations when users adjust wake routines. Schedule commands list, create, and update bedtime routines tied to the pod account. Audio subcommands expose state, play, and pause for integrated pod speakers. Base commands read info and set angles for adjustable mattress bases. The unofficial API is rate limited so agents should avoid repeated logins and confirm before changing temperature or alarms that affect sleep settings. Install via go install github.com/steipete/eightctl/cmd/eightctl@latest when the openclaw metadata reports the binary is absent. Install guidance covers go install when eightctl is absent. Operators should treat credential storage and rate limits as operational constraints when automating nightly temperature or schedule adj.3kinstalls44ImsgThe imsg skill wraps the imsg CLI for listing chats, reading history, watching new messages, and sending iMessage or SMS through macOS Messages.app. It applies only when the user explicitly asks for iMessage or SMS, not for Telegram, Signal, WhatsApp, Discord, or Slack channels that have dedicated tools. Requirements include macOS, Messages.app signed in, Full Disk Access for the terminal, and Automation permission for sending. Common commands cover imsg chats with JSON output, history by chat id with optional attachments, watch mode, and send with text, file attachments, and service flags for imessage, sms, or auto. Safety rules require confirming recipient and message content before send, avoiding unknown numbers without approval, verifying attachment paths, and self rate limiting to prevent spam. Bulk messaging always needs user confirmation first. The example workflow finds a contact chat, confirms the destination, then sends after approval.2.5kinstalls45Discorddiscord documents OpenClaw Discord operations through the shared message tool with channel set to discord. No separate Discord tool exists; all actions use JSON payloads respecting channels.discord.actions gates. Send targets use to channel or user IDs; mentions use user ID syntax. Common actions include send with silent flag, media attachments via file paths, components v2 for rich UI without mixing legacy embeds, react, read with limit, edit, delete, poll with duration hours and options, pin, thread-create with threadName, search across guild channels with query and channelIds, and set-presence with activityType when permitted. Rules require explicit guildId, channelId, messageId, and userId, multi-account accountId when needed, and avoiding Markdown tables in outbound messages. The skill requires channels.discord.token in config metadata and allows only the message tool for all Discord channel operations in the openclaw emoji metadata block.2.4kinstalls46WacliThe wacli skill documents the wacli CLI for third-party WhatsApp messaging and history sync, explicitly not for normal active user chats routed by OpenClaw. Safety rules require explicit recipient and message text with confirmation before send. Auth uses wacli auth QR login, wacli sync --follow for continuous sync, and wacli doctor for health checks. Chat discovery and message search support query filters, chat JIDs, and date ranges. History backfill requests additional messages when the phone stays online. Send covers text, group JIDs ending in @g.us, and file messages with captions. Storage defaults to ~/.wacli with optional --store override and --json for parsing. Install paths include Homebrew steipete/tap/wacli or go install from github.com/steipete/wacli. Agents should only reach for wacli when the user asks to message someone else on WhatsApp or to search synced history, never for routine conversation on the same WhatsApp thread.2.3kinstalls47Voice Callvoice-call integrates the OpenClaw voice-call plugin for initiating and inspecting phone calls through Twilio, Telnyx, Plivo, or mock providers. CLI examples use openclaw voicecall call with --to E.164 number and --message script, plus openclaw voicecall status --call-id for call state. The voice_call agent tool supports initiate_call with message and optional to and mode, continue_call and speak_to_user for mid-call messaging, end_call, and get_status actions. Configuration requires plugins.entries.voice-call.enabled and provider-specific credentials under plugins.entries.voice-call.config such as Twilio accountSid and authToken with fromNumber, Telnyx apiKey and connectionId, or Plivo authId and authToken. Mock provider enables development without live network calls. The skill is minimal and assumes the plugin is already configured before agents place outbound calls or check call status during conversational voice workflows on OpenClaw platforms.2.2kinstalls48SlackThe slack skill documents OpenClaw slack tool actions for channel messaging and member lookup. Inputs use channelId, messageId as Slack timestamps, to as channel or user targets, and emoji as Unicode or colon names. Actions include sendMessage, readMessages with limit, react, reactions list, editMessage, deleteMessage, pinMessage, unpinMessage, listPins, memberInfo, and emojiList, each as JSON action payloads. The agent should reuse channelId and message IDs from context when present rather than fuzzy name resolution. Safety rules require confirming destructive deletes when context is unclear, keeping outbound messages short, avoiding Markdown tables in Slack, and preferring thread or message IDs over fuzzy channel names. Configuration requires channels.slack in OpenClaw metadata. Teams invoke it when agents need to post updates, read channel history, add reactions, or manage pins in Slack workspaces connected to Clawdis.2.2kinstalls49Node Connectnode-connect is a steipete/clawdis skill that traces the real route from OpenClaw node to gateway across Android, iOS, and macOS clients. It forces topology classification first—same machine or emulator, same LAN, Tailscale tailnet, or public reverse proxy—before suggesting fixes, explicitly warning not to mix localhost debugging with VPS remote gateways. The skill verifies OpenClaw advertises the correct route, then resolves QR or setup code, pairing, and authentication failures. Developers reach for node-connect when node pairing hangs, setup codes fail, or connections drop across Wi-Fi versus Tailscale versus public URL paths.310installs50Autoreviewautoreview: A skill for development. This provides functionality for development workflows.268installs51Parallels Discord Roundtripparallels-discord-roundtrip: A skill for development. This provides functionality for development workflows.246installs52Openclaw Pr Maintaineropenclaw-pr-maintainer: A skill for development. This provides functionality for development workflows.238installs53Openclaw Ghsa Maintaineropenclaw-ghsa-maintainer: A skill for development. This provides functionality for development workflows.234installs54Openclaw Parallels Smokeopenclaw-parallels-smoke: A skill for development. This provides functionality for development workflows.234installs55Openclaw Test Heap Leaksopenclaw-test-heap-leaks: A skill for development. This provides functionality for development workflows.230installs56Security Triagesecurity-triage: A skill for development. This provides functionality for development workflows.219installs57Taskflowtaskflow: A skill for development. This provides functionality for development workflows.208installs58Taskflow Inbox Triagetaskflow-inbox-triage: A skill for development. This provides functionality for development workflows.174installs59Openclaw Qa Testingopenclaw-qa-testing: A skill for development. This provides functionality for development workflows.173installs60Openclaw Secret Scanning Maintaineropenclaw-secret-scanning-maintainer: A skill for development. This provides functionality for development workflows.159installs61Openclaw Test Performanceopenclaw-test-performance: A skill for development. This provides functionality for development workflows.144installs62Tag Duplicate Prs Issuestag-duplicate-prs-issues: A skill for development. This provides functionality for development workflows.135installs63Gitcrawlgitcrawl is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.126installs64Openclaw Testingopenclaw-testing is a Claude Code skill for testing & qa. It helps solo builders move faster with AI-assisted coding.125installs65Clawsweeperclawsweeper is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.120installs66Crabboxcrabbox is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.117installs67Discrawldiscrawl is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.115installs68Openclaw Small Bugfix Sweepopenclaw-small-bugfix-sweep is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.113installs69Discord Clawddiscord-clawd is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.112installs70Openclaw Debuggingopenclaw-debugging is a Claude Code skill for debugging. It helps solo builders move faster with AI-assisted coding.94installs71Diagram Makerdiagram-maker is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.92installs72Telegram Crabbox E2e Prooftelegram-crabbox-e2e-proof is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.88installs73Openclaw Refactor Docsopenclaw-refactor-docs is a Claude Code skill for code review & quality. It helps solo builders move faster with AI-assisted coding.87installs74Clawdtributorclawdtributor is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.77installs75Meme Makermeme-maker is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.76installs76Spikespike is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.73installs77Notcrawlnotcrawl is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.72installs78Slacrawlslacrawl is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.72installs79Channel Message Flowschannel-message-flows is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.71installs80Graincrawlgraincrawl is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.71installs81Openclaw Docker E2e Authoringopenclaw-docker-e2e-authoring is a Claude Code skill for security. It helps solo builders move faster with AI-assisted coding.70installs82Control Ui E2econtrol-ui-e2e is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.63installs83Openclaw Landable Bug Sweepopenclaw-landable-bug-sweep is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.63installs84Agent Transcriptagent-transcript is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.60installs85Python Debugpypython-debugpy is a Claude Code skill for debugging. It helps solo builders move faster with AI-assisted coding.60installs86Node Inspect Debuggernode-inspect-debugger is a Claude Code skill for debugging. It helps solo builders move faster with AI-assisted coding.59installs87Release Openclaw Maintainerrelease-openclaw-maintainer is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.57installs88Openclaw Changelog Updateopenclaw-changelog-update is a Claude Code skill for git & pull requests. It helps solo builders move faster with AI-assisted coding.56installs89Release Openclaw Cirelease-openclaw-ci is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.56installs90Release Openclaw Macrelease-openclaw-mac is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.56installs91Release Openclaw Nightlyrelease-openclaw-nightly is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.56installs92Release Openclaw Plugin Testingrelease-openclaw-plugin-testing is a Claude Code skill for testing & qa. It helps solo builders move faster with AI-assisted coding.56installs93Discord User Postdiscord-user-post is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.55installs94Verify Releaseverify-release is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.54installs95Release Openclaw Announcementrelease-openclaw-announcement is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.53installs96Canvascanvas is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.46installs97Technical Documentationtechnical-documentation is a Claude Code skill for documentation. It helps solo builders move faster with AI-assisted coding.44installs98Feishu DocExposes a single feishu_doc tool with actions to read, write, append, and create Feishu documents plus fine-grained block operations for tables and images. A developer uses it to read or programmatically edit Feishu cloud docs from an agent.2installs99Feishu DriveProvides a feishu_drive tool to list, inspect, create, move, and delete files and folders in Feishu cloud storage across doc/docx/sheet/bitable and other types. A developer uses it to manage Feishu Drive files from an agent.2installs100Feishu PermProvides a feishu_perm tool to list, add, and remove collaborators on Feishu docs, sheets, folders, and wikis across multiple member and permission types. A developer uses it to programmatically share Feishu files and manage access levels.2installs101Feishu WikiProvides a feishu_wiki tool to list knowledge spaces and nodes, resolve a node's obj_token, and create/move/rename wiki pages. A developer uses it to navigate a Feishu wiki and hand off to feishu_doc for reading or editing page content.2installs102ProseLoads the OpenProse VM spec so the agent executes .prose programs, routing prose commands (run, compile, update) and spawning sub-agent sessions for multi-agent orchestration. A developer uses it to author and run reusable multi-agent workflow scripts.2installs